New Approach to Investigate the Quality of Parmigiano Reggiano Cheese: Coupling of Structure Virtualization to Thermal, Rheological and Fracture Properties

The Parmigiano Reggiano cheese is regulated by a Protected Designation of Origin (PDO) Disciplinary, which imposes achievement of requirements about product characteristics and production process, as well as its commercial quality and designation of origin. One of the Disciplinary features concerns the structure. According to the disciplinary, the typical structure of the cheese is defined as finely-granulose structure (“pasta finemente granulosa”) and it brittle fracture (“frattura a scaglie”); therefore, when cheese is fractured, it breaks into scale-like fragments. Nevertheless, the two related-structure requirements are specified in the disciplinary without a clear definition nor an objective method to evaluate them. The aim of this PhD project was to investigate the cheese structure, fundamental rheological and fracture properties with non-destructive methodology and identification of the functional relationships between composition, microstructure and thermo-rheological properties in a multi-scale level: from sub-micron to macroscopic level. A second aim was to find a definition to flake fracture and being able to quantitatively determine it, and therefore develop an objective method for assessing fracture behaviour as a quality requirement for Parmigiano Reggiano. With these aims, more than 1700 cheese samples were obtained from cheese wheels from different raw milk composition and with a ripening age ranging from 12-months to 72-months, with the assumption of having the most different structures as possible. Cheese samples were provided by Parmigiano-Reggiano Consortium and selected among different cheese factories: 12 ripening times with 12, 14, 16, 18, 24, 28, 30, 36, 46, 54, 60 and 72 months of ripening. Then, each cheese wheel was divided in two parts along the sagittal plane to obtaining two specular half cheese wheels: A and B. Each half was divided in 10 cloves. Then, each clove was cut to obtain samples of parallelepiped-shaped to be analyzed under three-point bending test. In the three-point bending test, since there are discontinuities in the material, a standardized incision was created in the sample, in order to concentrate the stresses applied on the tip of the notch and not on the internal discontinuities, and reduce their variability. Then, from the half obtained from the bending test, cubic-shaped samples (2cm size) were cut, in order to be analyzed under uniaxial compression and isothermal creep tests. Finally, round disk-shaped samples were obtained from the second half of the bending test (with 20mm of diameter and 2.5mm of height), in order to be subjected to thermo-rheological tests. Results suggested that the fracture mode is strictly related to the heterogeneity of cheese structure over a large range of scale. The applied stresses concentrate around the tips of sub-micron discontinuities, this latter arise from partially fused curd junctions originated during milk clotting and cutting; consequently new surfaces originate and propagate along with the interfaces between fat and protein matrices due to their different relaxation times. The extent of crack propagation within the cheese bulk is limited by the presence of micro-voids, as well as by plastic and viscous dissipative forces both decreasing during ripening of the cheese. Bending tests quantitatively described the extent of mode-I fracture. Creep and stress relaxation data were analysed to compare cheese samples by relaxation times. Finally, temperature and frequency sweep oscillatory tests provided quantitative data of both elastic recovery and viscous or plastic deformations. Differential scanning calorimetry (DSC) has been used to evaluate the thermal behaviour, at a temperature between -80 and 350°C, with a heating rate of 10°C/min. Cheese structure has been characterized with imaging techniques, by means of Electron Scanning Environmental Microscopy (ESEM) and X-ray computed tomography. The specimens were first fractured to analyze the fracture surfaces, in order to investigate the 2D and 3D-microstructure on a microscopic and sub-microscopic scale. The ESEM with different types of detectors (SSD-BSD, LFD and EDAX) gave us different types of information about the structure: phases making up the composition, morphology and chemical elements. ESEM has allowed us to evaluate the distribution of water, proteins, fats and air, and scale in which they extend (sub-microscopic, as well as macroscopic scale). The casein structure looks branched, moisturizes and incorporated a certain amount of still unmelted fat globules. The fat phase, no longer globular, which was formed during the cooking of the curd, appears smooth. In addition, we assessed the macroscopic structure through the analysis of the fracture surface images with the use of a video camera and Image J software. The investigation of the macroscopic, microscopic and sub-microscopic structure, gave us details on the surface characteristics. The structure of the fracture surfaces is characterized by the partial overlap and complementarity of the fracture surfaces and by a structural irregularity. We therefore hypothesized that the mechanism of origin and propagation of the fracture depends on the structural characteristics of the surface. The structure of the PR originates from the caseous granules that partially melt together during the cooking of the curd. The fracture of the PR originates near the discontinuities that correspond to the intergranule junctions, i.e. interface between two or more granules. The junctions represent a preferential point for the concentration of the stresses applied from the outside both for the presence of microcracks and for the distribution of salt crystals. Both represent points where the two surfaces will detach. The propagation of the fracture therefore occurs in two ways: with an inter-granular mechanism if the fracture occurs following the intergranular junctions; and with a trans-granular type mechanism. In the latter case, the mechanically applied effort is distributed at the interface of different phases: the protein and lipid phase. Proteins and lipids in fact have different relaxation times and for this reason, at the interface of the two phases, a tension is created that facilitates the detachment of the two surfaces. An industrial tomograph was used to virtualize the inner structure and reconstruct three-dimensional volumes. With industrial tomography, a porous structure was highlighted, which was originated due to the effect of microbial fermentation. Tomographic analysis has allowed us to virtually reconstruct the three-dimensional distribution of the pores, and calculate their volume. In the more mature PR there are smaller and more numerous pores, while in the younger PR there are larger pores. From the evaluation of the distribution of the pores within the volume, it was deduced that the PR has an elasto-plastic behaviour and that the different distribution of the pores affects the deformation mechanisms. This means that the older PR will be elastic and the younger more plastic, because the coalescence mechanisms are more dissipative. This research sets the basis for the optimisation of the making process of Parmigiano Reggiano, considering fracture requirements as a key quality driver in a reversed engineering approach